摘要
电流互感器(CT)由于饱和使得副边电流变形,进而导致保护与测量应用中的许多问题。为此,提出一种补偿CT饱和特性的方法,以改善其测量性能。所提算法基于具有2个可变参数的sigmoid函数,构建了新型柔性神经网络,以估算CT励磁电流。实时地将估算所得励磁电流与扭曲的副边测量电流相加,即得补偿后原边电流。在学习过程中,所建补偿器的各柔性神经元柔性地改变其形状以适应各自的角色,高度柔性特点增强了网络学习能力,不但可减少网络节点数,而且可减少迭代学习时间。仿真研究中,应用一个900:5A的CT测试所提出的补偿器,测试时考虑了CT原边电流不同直流分量、CT剩磁大小与CT负载特性的影响。仿真结果验证了所提补偿方法的高精度,而且不受CT负载特性、CT剩磁情况及原边电流直流成分的影响。
The distorted secondary current of current transformer （CT） due to the saturation often results in many problems in applications of protections and measurements. A compensation method of CT saturation characteristic is proposed to improve its measure performance. The proposed algorithm uses a new flexible neural network with the activation function of two parameters changeable to realize estimation of magnetizing current. The accurate primary current is calculated by adding the estimated magnetizing current to the measured secondary current in real time. The flexible neurons of the established compensator will flexibly change the shape of each unit to adapt its role in learning process, which greatly simplifies the network with fewer neurons and reduces iterative learning epochs. A 900：5A CT is applied in the simulated operation, all kinds of cases such as the different direct current component of CT primary current, different CT .remanent flux level, and load characteristic, are considered when testing the performance of the compensator. The simulation results verify the high precision of the proposed compensation method regardless of the waveform of CT primary current, remanent flux level, and load characteristic.＂
出处
《中国电机工程学报》
EI
CSCD
北大核心
2006年第16期150-156,共7页
Proceedings of the CSEE
基金
北京交通大学论文基金(PD288)资助
关键词
电流互感器
饱和
柔性神经网络
current transformers
saturation
flexible neural network
